Correcting nonlinearity and temperature influence of sensors through B-spline modeling

A major problem of almost all types of industrial sensors is their nonlinear characteristics. Coming from the sensors' inherent physical peculiarities, nonlinearity can cause severe systematic measurement inaccuracy. As a countermeasure, the numerical technique of B-spline interpolation has been adapted to the field of electrical measuring. A great benefit of this method is the ability to correct even more than sensor nonlinearity. With a simple extension of this technique, the elimination of systematic external influences, such as the effect of temperature, is possible up to a high degree. In experiments with a GMR-based current sensor and a low-cost FPGA as digital platform, a measurement accuracy of 0.25% for current values up to ± 100A in a temperature range from −20°C to 80°C had been achieved.